Contents x
    Studio Overview
    • 10 Sep 2025
    • Print
    • Dark
      Light
    • PDF
    Contents

    Studio Overview

    • Updated On 10 Sep 2025
    • Print
    • Dark
      Light
    • PDF
    Article summary

    This article provides a comprehensive set of tools for labeling various types of data, supporting a wide range of AI and machine learning tasks. These studios are designed to streamline the annotation process, improve efficiency, and ensure high-quality labeled data for model training and evaluation.

    Dataloop offers specialized annotation studios to handle different data modalities, ensuring flexibility and scalability across various industries and use cases.

    Dataloop Annotation Studios

    Dataloop offers a suite of specialized Annotation Studios designed for different data modalities, each optimized with the right tools, automation, and workflows.

    Image Annotation Studio

    • Supports bounding boxes, polygons, key points, cuboids, polylines, and semantic segmentation.

    • Allows classification, tagging, and hierarchical annotation for complex labeling needs.

    • Enables automation features like model-assisted annotation and active learning to speed up labeling.

    Instance Segmentation Studio

    • Focuses on precise object segmentation using polygons and pixel-wise masks.

    • Supports automatic segmentation models and interactive tools like smart annotation.

    • Ideal for tasks requiring fine-grained object differentiation in images.

    Video Annotation Studio

    • Provides frame-by-frame object tracking, interpolation, and temporal segmentation.

    • Supports event classification and action recognition for video AI models.

    • Includes smart tracking and pre-labeling features to optimize efficiency.

    Audio Annotation Studio

    • Designed for annotating speech, music, environmental sounds, and more.

    • Supports audio segmentation, speaker dialyzation, and multi-channel audio labeling.

    • Enables transcription and classification tasks, often used for NLP models.

    Text Annotation Studio

    • Supports text classification, named entity recognition (NER), relation extraction, and sentiment analysis.

    • Enables multi-language annotation, including token-level, sentence-level, and document-level tagging.

    • Integrates GenAI models for AI-assisted labeling to speed up NLP tasks.

    PDF Annotation Studio

    • Data Extraction: Annotators label key fields (like dates, totals, names) in PDFs such as invoices or contracts to convert unstructured content into structured data for use in applications.

    • Content Review: Teams highlight, comment, and collaborate directly on PDFs for tasks like legal reviews, policy updates, or editorial checks, streamlining feedback and approval workflows.

    • ML/NLP Training Data: Annotated PDFs serve as ground truth for training machine learning models in tasks like entity recognition, or document classification.

    LiDAR Annotation Studio

    • Supports 3D point cloud annotation with bounding boxes, segmentation, and classification.

    • Works with multi-sensor fusion (LiDAR + images) for autonomous vehicle applications.

    • Provides semi-automatic tools for annotation acceleration.

    RLHF (Reinforcement Learning from Human Feedback) Annotation Studio

    • Designed for human preference labeling to fine-tune AI models.

    • Supports comparative ranking, rating scales, and open-ended feedback.

    • Primarily used in GenAI model training, such as LLMs and chatbots.

    GIS Annotation Studio

    • Geospatial data labeling with polygon and bounding box tools.

    • Satellite and aerial image segmentation.

    GenAI Evaluation Studio

    • GenAI Evaluation Studio is a powerful, user-friendly tool for evaluating GenAI responses, provide feedbacks, etc.

    • Allows users to design, run, and analyze model evaluations using fully customizable forms and layouts.

    Dataset in Read-Only Mode

    During export, the dataset enters Read-Only Mode to prevent changes. A warning message will appear in all annotation studios if the opened item belongs to a dataset currently being exported. While locked:

    • Saving and modifications are disabled.

    • Auto-save is off to avoid errors.

    • Save and Status buttons are disabled.

    • Actions will trigger an error message.

    🔄 Use the Refresh button to check the latest status. Developer or Project Owner can click Unlock to unlock the dataset if needed. Learn more

    Custom Annotation Studios

    Dataloop allows customers to create their own custom annotation studio applications tailored to their specific requirements, with the privacy scope as Organization. There are two ways to use the custom applications:

    1. Install the Custom Application: The custom applications are available to install in the entire projects in the organization. Refer to the Marketplace → Applications and install to your project.

    2. Share the Custom Application Service: User can open and work in a Custom Annotation Studio from another project where already installed, provided certain requirements are met. This makes it possible to collaborate and reuse applications across projects within your organization, without duplicating work.

    Share Studios Across Projects

    You can open and work in a Custom Annotation Studio from another project, provided certain requirements are met. This makes it possible to collaborate and reuse applications across projects within your organization, without duplicating work.

    Note: Sharing across projects applies only to custom studios, not to Dataloop’s built-in studios.

    Full Flexibility with SDKs

    With custom studios, you have complete control over both functionality and UI. Using Dataloop’s Python SDK and JavaScript SDK, you can:

    • Design unique layouts and workflows

    • Add custom tools and interactions

    • Extend the studio to meet specialized use cases

    Requirements

    To open a Studio from another project:

    1. Custom Studio Applications: Only custom studios can be opened across projects (built-in studios are not supported).

    2. Same Organization – The application’s service must belong to the same organization as your current project.

    3. Service Bot Membership – The application’s service bot must be added as a member of your current project.

    4. User Access – You must be a member of the original project where the application is hosted.

    To make your studio application available in other projects, refer to the steps here.

    Access Cross-Project Applications

    • Dataset Browser – In the dataset’s Open With menu, applications from other projects that meet the requirements will appear in the list.

    • Recipe View – In the recipe’s Open With menu, these applications will also be listed for quick access.

    Role-Based Access

    If you have restricted permissions, such as the Annotator role, your access to cross-project Studios will follow the same role-based permission rules as within the original project.

    What's Next